pith. sign in

arxiv: 2606.12449 · v1 · pith:ILQDUBFBnew · submitted 2026-05-30 · 🧬 q-bio.NC

A quantum-like benchmark for context-sensitive associative memory with adaptive plasticity

Yashine H. Goolam Hossen , Lea Gassab , Travis J. A. Craddock This is my paper

Pith reviewed 2026-06-28 17:26 UTC · model grok-4.3

classification 🧬 q-bio.NC
keywords associative memoryquantum-like formalismadaptive plasticitycontext sensitivitybenchmarktemporal organizationrecall dynamicshomeostatic stabilization
0
0 comments X

The pith

A benchmark for associative memory finds no universal quantum-like advantage, favoring multi-objective evaluation instead.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper develops an order-sensitive adaptive-plasticity benchmark to compare a quantum-like associative-memory model against matched real-valued no-phase and Markov-rate controls under identical task schedules, perturbations, and plasticity settings. It first identifies a narrow, non-monotonic weak-support regime where fixed structure alone cannot sustain recall without plasticity, then shows that adaptive plasticity, especially homeostatic stabilization, produces most performance gains. The Markov-rate control often reaches higher raw recall, yet the quantum-like model more reliably maintains order sensitivity and stage-dependent organization. The central result is that model families separate more clearly on a profile combining recall, temporal organization, and context sensitivity than on any single recall score.

Core claim

The paper establishes through controlled factorial comparisons that the quantum-like model does not exhibit a universal advantage in recall performance. Under weak structural support and regulated plasticity, most recall improvement stems from adaptive plasticity mechanisms rather than background connectivity. The Markov-rate control frequently delivers stronger raw recall, but the quantum-like formalism more consistently preserves order sensitivity and stage-dependent organization. These outcomes indicate that model classes are better distinguished by a multi-objective profile of recall, temporal organization, and context sensitivity than by isolated recall metrics.

What carries the argument

The order-sensitive adaptive-plasticity benchmark for staged associative recall, which applies matched weak-support screening, conservative plasticity operating points, and paired controls to isolate formalism effects on recall dynamics.

If this is right

  • Adaptive plasticity, especially homeostatic stabilization, supplies the primary recall gains when structural support is weak.
  • The quantum-like model preserves order sensitivity and context-dependent organization more reliably than the Markov-rate control.
  • Weak structure by itself fails to enable recall in the absence of plasticity.
  • Model classes require evaluation on combined metrics of recall, temporal organization, and context sensitivity rather than recall alone.
  • The useful weak-support regime remains narrow and non-monotonic, requiring explicit screening before comparisons.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The benchmark framework could map performance signatures for additional model families beyond the three tested here.
  • Extending the benchmark to varied perturbation profiles might clarify how context sensitivity scales with task complexity.
  • The multi-objective approach offers a template for comparing formalisms in other domains that require both stability and plasticity.
  • Matching benchmark parameters to empirical neural data could test whether the observed plasticity requirements align with biological memory.

Load-bearing premise

The screened weak-support regime and chosen conservative plasticity settings produce representative, fair comparisons across the quantum-like model and its classical controls.

What would settle it

An experiment showing that the quantum-like model loses its edge in order preservation and stage organization on the multi-objective profile, or that altering the plasticity operating point reverses the relative rankings, would undermine the claimed distinctions.

read the original abstract

Learning and memory require a balance between plasticity and stability: synaptic connections must encode new information without collapsing, saturating, or erasing previously useful structure. Associative-memory models can appear to learn successfully when fixed background connectivity already carries part of the task, making it difficult to distinguish genuine recall dynamics from structural assistance. We test this issue using an order-sensitive adaptive-plasticity benchmark for staged associative recall. The benchmark compares a quantum-like associative-memory model with matched real-valued no-phase and Markov-rate controls under the same task schedule, perturbation profiles, weak-support conditions, and plasticity settings. Here, "quantum-like" refers to the modeling formalism, not to a biological claim about quantum computation. We first screen weak structural support and then fix a conservative operating point for factorial comparisons across model families and plasticity mechanisms. The useful weak-support regime is narrow and non-monotonic. Weak structure alone does not rescue recall in the no-plasticity ablation, whereas most useful recall gains arise from adaptive plasticity, especially homeostatic stabilization. The Markov-rate control often achieves stronger raw recall, but the quantum-like model more consistently preserves order sensitivity and stage-dependent organization. These results do not support a universal quantum-like advantage. Instead, they show that model classes are better distinguished by a multi-objective profile combining recall, temporal organization, and context sensitivity than by any single recall score. The benchmark therefore provides a controlled framework for studying context-sensitive memory dynamics under weak support, regulated plasticity, and matched classical comparison.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript introduces an order-sensitive adaptive-plasticity benchmark for staged associative recall that compares a quantum-like associative-memory model against matched real-valued no-phase and Markov-rate controls under identical task schedules, perturbation profiles, weak-support conditions, and plasticity settings. After screening for weak structural support and fixing a conservative operating point, the work reports that the useful weak-support regime is narrow and non-monotonic, that weak structure alone does not rescue recall without plasticity, that the Markov-rate control often yields stronger raw recall while the quantum-like model better preserves order sensitivity and stage-dependent organization, and that model classes are therefore better distinguished by a multi-objective profile (recall + temporal organization + context sensitivity) than by any single recall score. No universal quantum-like advantage is claimed.

Significance. If the screened regime and matched controls produce representative comparisons, the benchmark supplies a controlled framework for distinguishing associative-memory formalisms under weak support and regulated plasticity, underscoring that multi-objective evaluation is more informative than raw recall alone and that adaptive plasticity (especially homeostatic stabilization) drives most gains.

major comments (2)
  1. [Abstract] Abstract: the central claim that the multi-objective profile distinguishes model classes more reliably than raw recall depends on the screened weak-support regime and the fixed conservative plasticity operating point being representative; the abstract states that this regime is narrow and non-monotonic yet provides no quantitative characterization of the screening procedure or independent justification for the chosen operating point, leaving open the possibility that the observed order-sensitivity advantage is an artifact of that specific slice.
  2. [Abstract] Abstract: the fairness of the no-phase and Markov-rate controls as matched comparators is load-bearing for the multi-objective distinction, but the abstract supplies no details on how perturbations interact with phase versus rate variables or on exact matching of plasticity mechanisms, so residual mismatches could undermine the reported differential preservation of order sensitivity.
minor comments (1)
  1. [Abstract] Abstract: the manuscript correctly clarifies that 'quantum-like' refers only to the modeling formalism and not to a biological claim, which is helpful for scope.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments on the abstract. We address each point below and will revise the abstract to incorporate additional quantitative and methodological details while preserving its concise nature.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that the multi-objective profile distinguishes model classes more reliably than raw recall depends on the screened weak-support regime and the fixed conservative plasticity operating point being representative; the abstract states that this regime is narrow and non-monotonic yet provides no quantitative characterization of the screening procedure or independent justification for the chosen operating point, leaving open the possibility that the observed order-sensitivity advantage is an artifact of that specific slice.

    Authors: The full manuscript details the screening in the Methods, including the parameter ranges for weak structural support and the selection criteria for the conservative operating point (chosen to ensure stability across model families and perturbation profiles). The abstract summarizes the outcome as narrow and non-monotonic but omits quantitative metrics. We will revise the abstract to include a brief quantitative characterization, such as the fraction of screened conditions showing non-monotonic behavior and the rationale for the operating point based on cross-model robustness, thereby strengthening the claim that the multi-objective distinction is representative within the screened regime. revision: yes

  2. Referee: [Abstract] Abstract: the fairness of the no-phase and Markov-rate controls as matched comparators is load-bearing for the multi-objective distinction, but the abstract supplies no details on how perturbations interact with phase versus rate variables or on exact matching of plasticity mechanisms, so residual mismatches could undermine the reported differential preservation of order sensitivity.

    Authors: The manuscript implements identical perturbation profiles and plasticity update rules across models, with phase and rate variables matched in scale and timing to ensure comparability. We agree the abstract does not explicitly state these matching details. In revision we will add a concise clause to the abstract summarizing that perturbations and plasticity mechanisms are applied identically, with variables matched in magnitude, to clarify the fairness of the controls and support the reported differential preservation of order sensitivity. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical benchmark with no self-referential derivations

full rationale

The manuscript describes a controlled empirical benchmark that screens weak structural support, fixes a conservative plasticity operating point, and compares model families (quantum-like vs. no-phase and Markov-rate controls) on multi-objective metrics under matched task schedules. No equations, parameter fits, or predictions are presented that reduce to their own inputs by construction. No self-citations are invoked as load-bearing uniqueness theorems or ansatzes. The central claim—that model classes are distinguished by a profile of recall, temporal organization, and context sensitivity rather than raw recall—is an empirical observation from the comparisons, not a definitional or fitted tautology. The narrowness of the weak-support regime is explicitly noted as a methodological finding rather than smuggled in via prior self-work.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; no explicit free parameters, axioms, or invented entities are stated. The modeling formalism itself and the assumption of matched controls are implicit but not detailed.

pith-pipeline@v0.9.1-grok · 5806 in / 1145 out tokens · 19610 ms · 2026-06-28T17:26:17.133831+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

32 extracted references · 22 canonical work pages

  1. [1]

    The Organization of Behavior: A Neuropsychological Theory

    Hebb DO. The Organization of Behavior: A Neuropsychological Theory. New York: Psychology Press; 2002. 2002 reissue of the original 1949 work

    2002

  2. [2]

    Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path

    Bliss TVP, Lømo T. Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. The Journal of Physiology. 1973;232(2):331–356. https://doi.org/10.1113/ jphysiol.1973.sp010273

    1973

  3. [3]

    Synaptic modifications in cultured hippocampal neurons: dependence on spike timing, synaptic strength, and postsynaptic cell type

    Bi GQ, Poo MM. Synaptic modifications in cultured hippocampal neurons: dependence on spike timing, synaptic strength, and postsynaptic cell type. The Journal of Neuroscience. 1998;18(24):10464–10472. https://doi.org/10.1523/ JNEUROSCI.18-24-10464.1998. 21

    arXiv 1998

  4. [4]

    VanderWeele

    Hopfield JJ. Neural networks and physical systems with emergent collective com- putational abilities. Proceedings of the National Academy of Sciences of the United States of America. 1982;79(8):2554–2558. https://doi.org/10.1073/pnas. 79.8.2554

    work page doi:10.1073/pnas 1982

  5. [5]

    Theoretical Neuroscience: Computational and Mathemat- ical Modeling of Neural Systems

    Dayan P, Abbott LF. Theoretical Neuroscience: Computational and Mathemat- ical Modeling of Neural Systems. Cambridge, MA: MIT Press; 2001

    2001

  6. [6]

    A simplified neuron model as a principal component analyzer

    Oja E. A simplified neuron model as a principal component analyzer. Jour- nal of Mathematical Biology. 1982;15(3):267–273. https://doi.org/10.1007/ BF00275687

    1982

  7. [7]

    Homeostatic synaptic plasticity: local and global mechanisms for stabilizing neuronal function

    Turrigiano G. Homeostatic synaptic plasticity: local and global mechanisms for stabilizing neuronal function. Cold Spring Harbor Perspectives in Biology. 2012;4(1):a005736. https://doi.org/10.1101/cshperspect.a005736

    work page doi:10.1101/cshperspect.a005736 2012

  8. [8]

    Home- ostatic role of heterosynaptic plasticity: models and experiments

    Chistiakova M, Bannon NM, Chen JY, Bazhenov M, Volgushev M. Home- ostatic role of heterosynaptic plasticity: models and experiments. Frontiers in Computational Neuroscience. 2015;9:89. https://doi.org/10.3389/fncom.2015. 00089

    work page doi:10.3389/fncom.2015 2015

  9. [9]

    Heterosynaptic plasticity and the experience-dependent refinement of developing neuronal circuits

    Jenks KR, Tsimring K, Ip JPK, Zepeda JC, Sur M. Heterosynaptic plasticity and the experience-dependent refinement of developing neuronal circuits. Frontiers in Neural Circuits. 2021;15:803401. https://doi.org/10.3389/fncir.2021.803401

    work page doi:10.3389/fncir.2021.803401 2021

  10. [10]

    Experience-dependent structural synaptic plasticity in the mammalian brain

    Holtmaat A, Svoboda K. Experience-dependent structural synaptic plasticity in the mammalian brain. Nature Reviews Neuroscience. 2009;10(9):647–658. https: //doi.org/10.1038/nrn2699

    work page doi:10.1038/nrn2699 2009

  11. [11]

    Dynamic models of large-scale brain activity

    Breakspear M. Dynamic models of large-scale brain activity. Nature Neuro- science. 2017;20(3):340–352. https://doi.org/10.1038/nn.4497

    work page doi:10.1038/nn.4497 2017

  12. [12]

    Quantum Models of Cognition and Decision

    Busemeyer JR, Bruza PD. Quantum Models of Cognition and Decision. Cambridge: Cambridge University Press; 2012

    2012

  13. [13]

    Pothos and Jerome R

    Pothos EM, Busemeyer JR. Can quantum probability provide a new direction for cognitive modeling? Behavioral and Brain Sciences. 2013;36(3):255–274. https: //doi.org/10.1017/S0140525X12001525

    work page doi:10.1017/s0140525x12001525 2013

  14. [14]

    Context effects produced by question orders reveal quantum nature of human judgments

    Wang Z, Solloway T, Shiffrin RM, Busemeyer JR. Context effects produced by question orders reveal quantum nature of human judgments. Proceed- ings of the National Academy of Sciences of the United States of America. 2014;111(26):9431–9436. https://doi.org/10.1073/pnas.1407756111

    work page doi:10.1073/pnas.1407756111 2014

  15. [15]

    Quantum-like states on complex synchronized networks

    Scholes GD. Quantum-like states on complex synchronized networks. Proceed- ings of the Royal Society A: Mathematical, Physical and Engineering Sciences. 2024;480(2295):20240209. https://doi.org/10.1098/rspa.2024.0209. 22

    work page doi:10.1098/rspa.2024.0209 2024

  16. [16]

    Dynamics in an emergent quantum-like state space generated by a nonlinear classical network

    Scholes GD. Dynamics in an emergent quantum-like state space generated by a nonlinear classical network. Physica Scripta. 2026;101(5):055205. https://doi. org/10.1088/1402-4896/ae3d90

    work page doi:10.1088/1402-4896/ae3d90 2026

  17. [17]

    Quantum information with quantumlike bits

    Amati G, Scholes GD. Quantum information with quantumlike bits. Physical Review A. 2025;111(6):062203. https://doi.org/10.1103/PhysRevA.111.062203

    work page doi:10.1103/physreva.111.062203 2025

  18. [18]

    Encoding quantumlike information in classical synchro- nizing dynamics

    Amati G, Scholes GD. Encoding quantumlike information in classical synchro- nizing dynamics. Phys Rev A. 2025 Sep;112:032423. https://doi.org/10.1103/ nvdp-sr11

    2025

  19. [19]

    Wichmann

    Geirhos R, Jacobsen JH, Michaelis C, Zemel R, Brendel W, Bethge M, et al. Shortcut Learning in Deep Neural Networks. Nature Machine Intelligence. 2020;2(11):665–673. https://doi.org/10.1038/s42256-020-00257-z

    work page doi:10.1038/s42256-020-00257-z 2020

  20. [20]

    Model selection and overfitting

    Lever J, Krzywinski M, Altman N. Model selection and overfitting. Nature Methods. 2016;13(9):703–704. https://doi.org/10.1038/nmeth.3968

    work page doi:10.1038/nmeth.3968 2016

  21. [21]

    Essential guidelines for computational method benchmarking

    Weber LM, Saelens W, Cannoodt R, Soneson C, Hapfelmeier A, Gardner PP, et al. Essential guidelines for computational method benchmarking. Genome Biology. 2019;20(1):125. https://doi.org/10.1186/s13059-019-1738-8

    work page doi:10.1186/s13059-019-1738-8 2019

  22. [22]

    Better than classical? The subtle art of benchmarking quantum machine learning models

    Bowles J, Ahmed S, Schuld M. Better than classical? The subtle art of benchmarking quantum machine learning models. arXiv. 2024;arXiv:2403.07059. [quant-ph]

    arXiv 2024

  23. [23]

    On Spectral Clustering: Analysis and an algorithm

    Ng AY, Jordan MI, Weiss Y. On Spectral Clustering: Analysis and an algorithm. In: Advances in Neural Information Processing Systems 14; 2001. p. 849–856

    2001

  24. [24]

    Modularity and community structure in networks

    Newman MEJ. Modularity and community structure in networks. Proceed- ings of the National Academy of Sciences of the United States of America. 2006;103(23):8577–8582. https://doi.org/10.1073/pnas.0601602103

    work page doi:10.1073/pnas.0601602103 2006

  25. [25]

    Survey of multi-objective optimization methods for engi- neering

    Marler RT, Arora JS. Survey of multi-objective optimization methods for engi- neering. Structural and Multidisciplinary Optimization. 2004;26(6):369–395. https://doi.org/10.1007/s00158-003-0368-6

    work page doi:10.1007/s00158-003-0368-6 2004

  26. [26]

    Meyarivan

    Deb K, Pratap A, Agarwal S, Meyarivan T. A fast and elitist multiobjective genetic algorithm: NSGA-II. IEEE Transactions on Evolutionary Computation. 2002;6(2):182–197. https://doi.org/10.1109/4235.996017

    work page doi:10.1109/4235.996017 2002

  27. [27]

    Behavioral timescale synaptic plasticity: properties, elements and functions

    Magee JC. Behavioral timescale synaptic plasticity: properties, elements and functions. Nature Neuroscience. 2026;29:1–15. https://doi.org/10.1038/ s41593-026-02214-2

    2026

  28. [28]

    Cutts and Stephen J

    Madar AD, Milstein AD, O’Dell TJ, Jain A, Clopath C, Sheffield MEJ. Behavioral Timescale Synaptic Plasticity: A Burst in the Field of Learning and Memory. 23 The Journal of Neuroscience. 2025;45(46). https://doi.org/10.1523/JNEUROSCI. 1332-25.2025

    work page doi:10.1523/jneurosci 2025

  29. [29]

    The temporal paradox of Hebbian learning and homeostatic plasticity

    Zenke F, Gerstner W, Ganguli S. The temporal paradox of Hebbian learning and homeostatic plasticity. Current Opinion in Neurobiology. 2017;43:166–176. https://doi.org/10.1016/j.conb.2017.03.015

    work page doi:10.1016/j.conb.2017.03.015 2017

  30. [30]

    Modeling the dynamic interaction of Hebbian and homeostatic plasticity

    Toyoizumi T, Kaneko M, Stryker MP, Miller KD. Modeling the dynamic interaction of Hebbian and homeostatic plasticity. Neuron. 2014;84(2):497–510. https://doi.org/10.1016/j.neuron.2014.09.036

    work page doi:10.1016/j.neuron.2014.09.036 2014

  31. [31]

    jbusres.2019.07.039 Solomonides, A

    Butz M, W¨ org¨ otter F, van Ooyen A. Activity-dependent structural plastic- ity. Brain Research Reviews. 2009;60(2):287–305. https://doi.org/10.1016/j. brainresrev.2008.12.023

    work page doi:10.1016/j 2009

  32. [32]

    Memory preser- vation and cooperative shielding in complex quantum networks

    Ausilio S, Borgonovi F, Celardo GL, Yago Malo J, Chiofalo ML. Memory preser- vation and cooperative shielding in complex quantum networks. Physical Review B. 2026;113(6):064201. https://doi.org/10.1103/dt4n-8sqg. 24 Supplementary Information S1 Scope of this Supplementary Information This Supplementary Information provides the technical context underlying...

    work page doi:10.1103/dt4n-8sqg 2026